Show HN: Trelk – Read, Think, Connect
The article discusses Trelk, a platform that utilizes Google's state-of-the-art embedding model, EmbeddingGemma 300M. This model is designed for on-device deployment and supports over 100 languages, making it a versatile tool for various applications. The model's capabilities include semantic search and clustering, allowing for efficient data processing and analysis.
- ▪Trelk uses Google's EmbeddingGemma 300M model, which has 300M parameters and 768-dimension vectors.
- ▪The model supports multilingual processing for over 100 languages.
- ▪The EmbeddingGemma 300M model is designed for on-device deployment, enabling efficient processing and analysis.
Opening excerpt (first ~120 words) tap to expand
Embeddings EmbeddingGemma 300M Google's state-of-the-art embedding model built from Gemma 3 and the same research behind Gemini. 300M parameters, 768-dimension vectors, 2K token context. Multilingual support for 100+ languages. Designed specifically for on-device deployment. Semantic search Clustering 768 dimensions On-device
Excerpt limited to ~120 words for fair-use compliance. The full article is at Trelk.
Discussion
0 commentsMore from Trelk
